Vitis Vinifera) Jason P

Londo et al. Horticulture Research (2018) 5:10 DOI 10.1038/s41438-018-0020-7 Horticulture Research

ARTICLE Open Access Divergence in the transcriptional landscape between low temperature and freeze shock in cultivated grapevine (Vitis vinifera) Jason P. Londo1,2, Alisson P. Kovaleski2 and Jacquelyn A. Lillis1,3

Abstract Low-temperature stresses limit the sustainability and productivity of grapevines when early spring frosts damage young grapevine leaves. Spring conditions often expose grapevines to low, but not damaging, chilling temperatures and these temperatures have been shown to increase freeze resistance in other model systems. In this study, we examined whole-transcriptome gene expression patterns of young leaf tissue from cuttings of five different grapevine cultivars, exposed to chill and freeze shock, in order to understand the underlying transcriptional landscape associated with cold stress response. No visible damage was observed when grapevine leaves were exposed to chilling temperatures while freeze temperatures resulted in variable damage in all cultivars. Significant differences in gene expression were observed between warm control conditions and all types of cold stress. Exposure to chill stress (4 °C) versus freezing stress (−3 °C) resulted in very different patterns of gene expression and enriched pathway responses. Genes from the ethylene signaling, ABA signaling, the AP2/ERF, WRKY, and NAC transcription factor families, and starch/sucrose/galactose pathways were among the most commonly observed to be differentially regulated. Preconditioning leaves to chill temperatures prior to freezing temperatures resulted in slight buffering of gene expression responses, suggesting that differences between chill and freeze shock perception complicates 1234567890():,; 1234567890():,; identification of candidate genes for cold resistance in grapevine. Overall, the transcriptional landscape contrasts observed between low temperature and freezing stresses demonstrate very different activation of candidate pathways impacting grapevine cold response.

Introduction restricting production to cultivars that can survive winter Environmental stresses are the leading factors limiting or through production of hybrid varieties. In addition, the current and future expansion of grapevine production increased variation in fall and spring weather patterns can in the United States. As a Mediterranean-adapted species, lead to frost and freeze damage as vines are preparing for cultivated grapevine (Vitis vinifera) is best suited to cli- winter in autumn, and also in spring, as buds exit dor- mates with mild winter conditions and with relatively mancy and tender new growth is present. Frost risk is a consistent temperature change in autumn and spring. constant concern for fruit crop production as young green However, grapevine is often grown outside of this climatic leaves lack the physical defenses associated with freeze niche in regions that have severe winter conditions, resistance during dormancy (freeze barriers and dehy- – drated tissues)1 3. There is no indication that winter variability will cease, thus there is a great need for better Correspondence: Jason P. Londo ([email protected]) 1United States Department of Agriculture, Agricultural Research Service, Grape understanding of grapevine chill and freeze exposure Genetics Research Unit, 630 W. North Street, Geneva, NY, USA response in order to identify candidate genes for future 2 School of Integrative Plant Science, Horticulture section, Cornell University- breeding of hybrid varieties suited for these challenging New York State Agricultural Experiment Station, 630 W. North Street, Geneva, NY, USA growing conditions. Full list of author information is available at the end of the article

© The Author(s) 2018 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a linktotheCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Londo et al. Horticulture Research (2018) 5:10 Page 2 of 14

Studies of freeze shock (<0 °C) and chill shock (4–10 °C) As a result of these studies, different grapevine cultivars stresses in a number of different crop and plant species and grapevine species have been ranked as cold hardy have resulted in a common cold response (COR) gene or cold tender. The majority of transcriptomic studies regulation and signaling cascade. While a temperature examining cold-temperature effects in grapevine involve – sensor has yet to be confirmed, plants are able to perceive studies of leaf or berry tissue19 23. During these studies, low temperature and initiate a gene regulation cascade cold stress is often applied as low, non-freezing that prepares the plant to tolerate or resist freeze damage. temperatures such as 4 °C. These studies have demon- This cascade begins with the activation of calcium sig- strated a interconnected web of gene regulation shared naling, triggering downstream MAPK kinase activity4, 5, with aspects of drought and salt stresses as well as hor- and changes in the activity of the BHLH transcription mone responses and biotic stress10. Cold stress results in factor family ICE. These ICE genes, in turn, bind to DNA the activation of hundreds of genes with the vast majority and activate the expression of the cold-binding factor/ of their function related to changes in transcription fac- dehydration responsive element-binding (CBF/DREB) tors and activation of genes with poorly annotated – transcription factors6 8. CBF/DREB genes in turn bind to information. These studies are important for under- the DNA and activate many other downstream genes, standing the transcriptional response and potential miti- called COR genes, leading to changes in cryoprotectant gation pathways associated with cold stress damage, but levels and sugar metabolism, as well as impacts on growth they have not been conducted under freezing conditions. and cell expansion9, 10. The elements of this CBF response Further complicating the issue, comparisons of cold have been observed across the plant kingdom and several “hardy” and “sensitive” genotypes may not reflect actual – of the key genes have been identified in grapevine11 13. leaf freezing response as these phenotypic designations Many plants go through a process called acclimation are due to the resistance of dormant tissues, not green when they are exposed to low, but non-freezing tem- growing tissues. peratures. This process is believed to contribute to Studies have been conducted to examine specific gene enhanced freezing resistance by giving the plant time to families (e.g., WRKY) involved in cold stress response24 or adjust its physiology to resist the stresses of ice formation, on whole-transcriptome analysis of low-temperature such as changes in membrane structure, alterations in stress in grapevine comparing cultivated and wild spe- sugar concentrations, and production of cryoproteins14. cies21, but no analysis of the transcriptomic landscape has Studies in Arabidopsis have demonstrated an enhance- been conducted to elucidate the potential differences ment of freeze tolerance and survival after exposing plants between low temperature and freeze stresses in grapevine. to an acclimating temperature of 4 °C for 24 h. However, a Therefore, the objective of this study was to identify previous study in grapevine demonstrated that pretreat- transcripts and pathways that are differentially expressed ment of swollen buds with the same acclimating tem- in young leaf tissue, and enriched in response to various perature was not effective at reducing freeze damage15. cold stresses. We test the hypothesis that exposure to low Few other studies have been conducted to examine the temperature increases freeze resistance in grapevine using potential resistance of grapevine growing tissues to cold contrasts between temperature exposures to explore the temperatures or to examine variation in freeze response of primary transcriptomic processes associated with accli- leaves. The leaves of grapevine are not considered parti- mation/chill (4 °C) exposure, freeze shock (−3 °C), and cularly freeze resistant, typically suffering damage at freeze shock after acclimation (4 to −3 °C). Pathway temperatures below −2.5 °C15. Studies examining frost enrichment analysis was applied to this data to identify protectants have demonstrated that leaves can survive the significant changes in gene expression that are shared these low damaging temperatures if ice nucleation is among and differ between V. vinifera cultivars. prevented through the use of kaolin clay or through eliminating ice-nucleating bacteria16. These studies have Materials and Methods demonstrated that leaves can survive freezing tempera- Plant Material and Treatments tures “by chance” and sporadic damage in vineyards fol- Five different V. vinifera cultivars (“Riesling”, “Cabernet lowing a frost are likely due to stochastic avoidance of ice Franc”, “Chardonnay”, “Sangiovese”, and “Tocai Fruliano”) formation and leaf supercooling. It remains unknown were collected as dormant canes from the field in mid- whether actual freeze resistance mechanisms and varia- winter (December 2013) and cut into single node cuttings. tion between cultivars for leaf freeze resistance occurs. In Cuttings were held in a cold room for 1000 h (~42 days at contrast to the lack of studies in leaf freeze resistance, 4 °C) in order to synchronize dormancy state25. Cuttings many studies have been performed examining the cold were then transferred to a growth room with a tempera- hardiness of grapevine dormant tissues using methods ture of 22 °C and a light cycle of 16/8 light/dark to initiate such as differential thermal analysis to determine lethal growth and bud burst. When cuttings had reached the temperatures for bud and cane tissues during winter17, 18. E–L stage of 726, cuttings were randomized and placed Londo et al. Horticulture Research (2018) 5:10 Page 3 of 14

into styrofoam trays with the basal end encased in par- reverse replicate_cultivar_treatment.bam V1.gff). Nor- afilm wax and treated with one of four conditions: malization and differential expression were completed 1. Control, maintained in growth chamber until using the edgeR package following the default workflow collection of tissue for RNA. with the following exceptions. Library adjusted counts 2. Acclimated/Chill stress (hereafter: acclimation/ per million (CPM) values were computed and lowly acclimated), placed in 4 °C for 48 h, then collected expressed genes were excluded by removing features for RNA. with CPM <2 and without expression in at least 12 3. Acclimated freeze, placed in 4 °C for 48 h, then libraries for analyses with cultivars combined, and at exposed to a programed temperature cycle of 4 °C least 3 libraries for within cultivar analysis (Supple- for 30 min, decreasing temperature ramp 0.5°/min to mental Table 2). Differentially expressed (DE)genes −3.0 °C, hold for 45 min at −3 °C, increasing were defined based on a FDR corrected p-value <0.05 temperature ramp 0.5 °C/min to 4 °C, hold for 30 and at least a 2-fold change (Supplemental Data 1). min. Tissues were collected for RNAseq at the end Illumina sequence data are available in the sequence of the 45 min hold at −3.0 °C. reads archive under the Bioproject number SUB3019356 4. Non-Acclimated-Freeze (Freeze), no 4 °C pre- with accession numbers SAMN07617826- treatment, but the same programed freeze exposure SAMN07617885. outlined above. Cuttings were lightly misted with dH2O prior to being Pathway and Enrichment Analysis placed in the freezer chamber in order to ensure that DE gene lists were compiled for treatment contrasts, uniform freeze occurred across all samples in the chamber pooling all cultivars, to produce data sets of up and and to prevent random leaf supercooling. Six replicate downregulated genes for each treatment. DE lists for each freeze stress runs were performed for the control freeze cultivar were also compiled at the treatment level. After and acclimation freeze tests and visual damage assess- determining DE genes, the program VitisPathways28 ments were performed 4 days after freeze exposure. (http://momtong.rit.edu/cgi-bin/VitisPathways/ vitispathways2.cgi), was used to identify biological path- RNAseq Libraries and Quality Control ways and molecular functions being affected by treatment, Leaf tissue was collected in triplicate from E–L 7 stage leveraging the Vitisnet database (https://www.sdstate.edu/ leaves in each of the treatment conditions and immedi- vitisnet-molecular-networks-grapevine)29. Enriched path- ately frozen with liquid nitrogen. For tissue collected from ways of DE genes were determined using 100 permuta- freeze treatments, only green tissue without obvious tions, a Fisher’s exact test of p < 0.05, and a permuted p- freeze damage (large discolored sectors) was collected to value of <0.1. Pathways that were enriched both in the avoid biasing sampling toward cell death. Total RNA was combined analysis as well as observed in cultivar × extracted from these tissues using Sigma Spectrum kits treatment analyses, present in at least four of five culti- (Sigma-Aldrich, St. Louis, MO, USA) and strand-specific vars, were determined to represent consistent and libraries prepared27. Three replicate samples were necessary stress responsive pathways. These pathways extracted and prepared for each cultivar × treatment were further investigated to elucidate the specific genes combination for a total of five cultivars, four treatments, which are differentially regulated under these conditions. and three replicates. These libraries were multiplexed Shared and unique DE genes and enriched pathways with barcode adapters and sequenced as 100 bp single- were visualized with InteractiVenn30 (https://omictools. end reads on an Illumina 2500. com/interactivenn-tool). Candidate pathways and logFC Raw reads were demultiplexed based on custom 6 bp results for genes within these pathways were visualized barcodes. Residual adapter and low quality sequences using Cytoscape 3.4.0 (http://www.cytoscape.org/). were removed using Fastx-tools (fastq_quality_filter –Q33 –q25 –p 25; fastx_trimmer –Q33 –f7;cutadapt Results rcprAC rcprBC –minimum-len 25 –O 3). Final read Phenotypic Response of Grapevine Leaves quality was evaluated using FastQC. Quality reads were No apparent damage to leaves were noted in control or aligned to the V. vinifera 12xV2 genome (https:// acclimated treatment. Following the two freeze treatment urgi.versailles.inra.fr/Species/Vitis)) and the V1 anno- conditions, damage was initially noted as leaf tissue with a tation (http://genomes.cribi.unipd.it/grape/)usingwatery appearance, likely indicating tissue damage and Tophat (tophat -p 48 --library-type = fr-firststrand -o cell leakage. After 4 days of recovery, these damaged outputfilename -G V1.gff replicate_cultivar_treatment. regions had dried out and appeared as large lesions. A fq). Alignment statistics can be found in Supplemental wide range in damage was observed in freeze-exposed Table 1. Uniquely aligned reads were quantified for all leaves, ranging from the least-damaged cultivar “San- V1 annotated genes using HTSeq (htseq-count -f bam -s giovese”, with ~10% damage, to the most-damaged Londo et al. Horticulture Research (2018) 5:10 Page 4 of 14

“Chardonnay”, with ~60% damage. All five cultivars in this Table 1). This library, replicate 1 for Riesling in warm study had increased damage from freezing following conditions, was removed from the analysis. A total of 18 acclimation treatments. “Sangiovese” suffered only slightly 367 genes were detected in this study after filtering for more damage relative to control test but “Cabernet Franc” low expression values and when analyzed with all cultivars experienced a dramatic increase in freeze damage from grouped within treatment, representing 61% of the 29 971 ~32% to ~92% after acclimation period (Fig. 1). genes represented in the Cribi-V.1 gene annotation (http://genomes.cribi.unipd.it/grape/). An MDS plot was RNAseq Data Quality and Read Count Statistics used to visually assess the effects of cultivar and treatment The mean size of the sample libraries was 4.6 M reads on global gene expression patterns. Treatments tended to after QC and alignment. Spearman's correlations were cluster separately with overlap between acclimation and examined for cultivar libraries and a single library was acclimation freeze treatments while differences driven by determined to represent an outlier, likely due to insuffi- cultivar were not apparent in the represented axes (Sup- cient sequenced reads (Supplemental Fig. 1; Supplemental plemental Fig. 2).

Differential Expression Contrasts between warm and stress treatments revealed 8070 genes that were differentially expressed at LogFC >1 and FDR <0.05. In general, stress treatments resulted in a greater number of genes that were downregulated rather than were upregulated. Hierarchical clustering of samples using differentially regulated genes are able to delineate treatment affects. Acclimation freeze samples are found to cluster both with freeze and with acclimation treatments, demonstrating overlap between responsive genes in these treatments (Supplemental Fig. 3). Acclimation stress resulted in the least number of DE genes and freeze treatments resulted in the greatest number. Stress treatments resulted in 214 shared upregulated genes and Fig. 1 Phenotypic response observed on leaf tissue following 884 shared downregulated genes. For upregulated genes, freeze treatments. Bars are the average of six replicated experiments, acclimation and acclimation freeze shared 414 upregu- error bars represent standard error of the mean lated genes and 631 downregulated genes while

Fig. 2 Differentially expressed genes based on treatment. Venn diagram of shared upregulated and downregulated DE genes in acclimation and freeze stress treatments Londo et al. Horticulture Research (2018) 5:10 Page 5 of 14

acclimation freeze and freeze shared 781 genes and 743 Comparing the effects of stress, averaged across cultivars, genes respectively. Only 5 genes were observed to be and by also checking for enrichment within cultivar, upregulated in both acclimation and freeze treatments allowed us to identify candidate pathways that were and 49 were shared as downregulated (Fig. 2). Lists of consistently reactive to stress treatments. Of the many shared DE genes grouped by treatment with functional enriched pathways detected within this experiment, we annotation can be found in Supplemental Data 2. Counts identiﬁed several key pathways that were repeatedly upre- of DE genes for treatment averages and for cultivar spe- gulated and represented among comparisons of stress ciﬁc contrasts demonstrate a similar pattern of high DE treatments and shared among cultivars. (Supplemental genes for freeze treatment comparisons and low DE genes Data 3 and 4). These enriched pathways are described for acclimation treatment (Supplemental Table 3). below. LogFC information and Vitis gene loci represented in these pathways can be found in Supplemental Data 5-10. Pathway Enrichment Results of the pathway analysis demonstrated that there Ethylene were an abundance of enriched pathways in the process The transcriptional landscape of the ethylene pathway, categories of metabolism, environmental processing, and including ethylene synthesis, signaling, and downstream transcription factors, and a reduced enrichment in processes transcription factor activation was very different between of transcription, cellular processes and transport (Table 1). acclimation stress and freeze stresses (Fig. 3). Genes in the Further investigation within these stress-responsive ethylene production pathway were upregulated in freeze categories revealed overrepresentation of subcategories of carbohydrate, lipid, and amino acid metabolism, as well as biosynthesis of secondary metabolites and pathways associated with hormone signaling. A large number of DE genes were enriched in different transcription factor families, within both up and downregulated responses.

Table 1 Sample of the pathways and processes with signiﬁcant enrichment of DE genes, partitioned for up and downregulated genes in stress

Pathway or process Acclimation Freeze Acclimation freeze

Down Up Down Up Down Up

1. Metabolism 11 11 9 29 17 13 1.1 Carbohydrate 221622 metabolism 1.3 Lipid metabolism 0 42432 1.5 Amino acid 131714 metabolism 1.9 Biosynthesis of 510724 secondary metabolites 2. Transcription 2 22020 3. Environmental 432435 processing 3.2 Hormone 322233 signaling 4. Cellular processes 3 12030 Fig. 3 Reduced Vitisnet ethylene synthesis and signaling pathway demonstrating log fold change in differentially 5. Transport 2 43543 expressed genes as a result of stress treatment. Each box in a 6. Transcription factors 12 8 15 15 11 18 series represents a cribi-V1 annotated paralog for the indicated gene. Genes without boxes were not differentially expressed in this Pathways and processes follow VitisNet designations. Numbers in columns indicate experiment the number of pathways within the classiﬁcation that were identiﬁed as enriched. Londo et al. Horticulture Research (2018) 5:10 Page 6 of 14

Fig. 4 Reduced Vitisnet ABA synthesis and signaling pathway demonstrating log fold change in differentially expressed genes as a result of stress treatment. Each box in a series represents a cribi-V1 annotated paralog for the indicated gene. Genes without boxes were not differentially expressed in this experiment. Blue arrows indicate patterns of differential expression associated with freeze stress. Genes that repress ABA synthesis shown in red boxes. ABA-dependent transcription factors in black boxes and acclimation freeze treatments, suggesting that ethy- particularly in the downstream signaling aspects leading lene levels may increase in the leaves. Negative regulators to ABA-based transcription factor activation. Since ethy- of ethylene signaling downstream of ethylene synthesis lene signaling was not seen to be repressed (Fig. 3), the were not upregulated. interaction between ABA and ethylene seems to be a coordinated, rather than antagonistic interaction leading Abscisic Acid (ABA) to large changes in downstream stress responsive tran- Enrichment of three pathways was consistently scription factors. observed across temperature treatments that represent components of abscisic acid (ABA): carotenoid bio- NAC, WRKY, AP2/ERF synthesis (secondary products), ABA biosynthesis, and Several transcription factor families were signiﬁcantly ABA signaling (Fig. 4). The effect of freeze stress is enriched in this experiment including the HSF, PLATZ, strongly seen in the expression of NCED3, the commit- MYB, GRAS, NAC, WRKY, and AP2/ERF families. Of ting step in ABA synthesis31. Clear differences were these, the NAC, WRKY, and AP2/ERF families were observed between acclimation and freeze stresses, speciﬁcally enriched for genes that were upregulated in Londo et al. Horticulture Research (2018) 5:10 Page 7 of 14

Fig. 5 NAC, WRKY, and AP2/ERF transcription factor families demonstrating log fold change in differentially expressed genes as a result of stress treatment. Each box in a series represents a cribi-V1 annotated gene. Composite image compiled to demonstrate shared, contrasting, and unique responses of TF genes stress treatments (Fig. 5). For the NAC and WRKY of upregulated genes in these gene families and acclima- families, the signature of stress response was very similar. tion freeze responses were typically lower in fold change In acclimation stress, there were a few DE, upregulated than that of freeze stress. genes, but a greater number of downregulated genes. In Major changes in the AP2/ERF gene family were contrast, freeze stress resulted in a much greater number observed between acclimation and freeze treatments with Londo et al. Horticulture Research (2018) 5:10 Page 8 of 14

acclimation freeze resulting in a similar, but muted pat- determine if any pathways were overrepresented and tern relative to freeze shock. The composite portion of might be linked to increased freeze resistance. 567 DE Fig. 5 demonstrates the genes which were related to genes were unique to “Sangiovese” and these genes were specific contrasts in this experiment; “Cold” (same logFC shown to result in 16 significantly enriched pathways in pattern in all cold treatments), “Chill” (associated with Vitis (Supplemental data 11). Of particular interest is the acclimation treatment), “Freeze” (shared regulation in phenylpropanoid biosynthesis pathway which had “San- both freeze treatments), “Shock” (only associated with giovese” specific transcripts for phenylalanine ammonia freeze shock), and “Contrast” (regulation was in opposite lyase and stilbene synthase, the committed steps leading directions in stress). The relative proportion within these to the production of stilbenes, secondary products asso- categories indicate that shared responses only represent ciated with stress response. Of the 42 predicted stilbene around 6–8% of the observed gene expression changes synthase genes in the Vitis annotation, 24 were upregu- with “Freeze”, “Shock”, and “Contrast” each representing lated within freeze and 15 were upregulated within the dominant patterns of expression. acclimation freeze treatments in “Sangiovese”.

Sugar Metabolism Discussion The pathways associated with sugar metabolism, spe- Several different studies have been conducted looking at cifically the starch, sucrose, and galactose pathways, were the effects of low temperature on the grapevine leaf significantly enriched for DE genes. When examining the transcriptome. Transcriptome differences have been interaction of these complex pathways (Fig. 6)we evaluated in comparisons between low temperature and observed that all stress treatments resulted in major other abiotic stresses like drought or salt19, between 22, 24 downregulation of genes leading to the production of D- species with varied cold hardiness phenotypes ,or 21 Glucose (Q and S), D-Glucose 6-Phospahte (D, E, and V), looking at specific gene families . In each of these studies, D-Fructose 6-Phosphate (D, E, and G), as well as the the applied cold stress is that of a low temperature, about degradation of Lactose (J) and Trehalose (A and EE). All 4–5 °C. This is the same temperature used as an accli- stress treatments also seemed to trigger enzymes mation temperature in other crop and model systems to responsible for production of Amylose, Starch, Maltose, prime a plant to develop increased freeze hardiness. Since and Dextrin (W, X, Y, R, AA, BB, CC, and DD). Specific most designations of hardiness in grapevine describes the differences were observed in the response of acclimation freeze resistance of dormant tissues such as buds and versus the two freeze stress treatments. These treatments canes, comparisons of hardy versus sensitive grape vari- resulted in differing expression patterns for genes eties at low, non-freezing temperatures in leaf tissue may involved with the conversion of UDP-galactose to α-D- not completely elucidate the important genes and path- galactosyl-1,3-1D-myo-inositol, a precursor to raffinose, ways for freeze resistance in grape. In this study we and in the conversion of UDP-glucose to D-glucoside, a wanted to better understand how the transcriptomes of a common precursor that enters into many secondary leaf exposed to low temperature compared with that of a product biosynthesis pathways. The transcriptional land- leaf exposed to simulated frost/freeze conditions to scape of these sugar pathways thus demonstrate a shunt determine whether acclimation of growing tissues leads to away from growth-based metabolic substrates like glucose enhanced freeze resistance. In addition, we wanted to lay and fructose in control leaves, toward that of storage the foundation of the grapevine freeze stress tran- (starch) in chill stressed leaves, and toward storage, scriptome for future studies of dormant tissue response. osmotic protection (raffinose), and defense response (secondary products) in freeze stress leaves. Phenotypic Response to Stress Treatments In this study, cuttings were exposed to warm control Phenotype and Pathway Integration conditions at 22 °C until they reached the E–L stage 7, a When examining the phenotype of freeze resistance, stage where field damage due to frost is commonly many differences between the five cultivars studied could observed. No leaf damage was observed in these control be observed. A list of the top upregulated and down- conditions, nor on cuttings exposed to low temperature regulated genes as well as those genes found consistently (4 °C). This was in stark contrast with cuttings treated is reported in Supplemental Data 3. Of particular interest with a simulated frost/freeze event which showed varied in this study was “Sangiovese”, which appeared to resist evidence of freeze damage as large sectors of leaves freeze damage both with and without acclimation treat- appearing to be dark in color and water saturated. Pre- ment. We examined all upregulated genes across the five sumably, these sections were parts of the leaf where cell cultivars in both freeze and acclimation freeze treatments membranes had ruptured and cellular fluids had saturated by cultivar and identified the genes which were uniquely the tissues. When cuttings were first acclimated at 4 °C for differentially expressed in “Sangiovese” in order to 48 h and then treated with the simulated frost/freeze, Londo et al. Horticulture Research (2018) 5:10 Page 9 of 14

Fig. 6 Sucrose, starch, and galactose pathways demonstrating major shifts in the expression of metabolic enzymes in response to chill and freeze stresses. Metabolic enzymes denoted with letters for brevity (supplemental data 10). Blue indicates signiﬁcant fold upregulation, brown indicates signiﬁcant downregulation. Blue directional demonstrate presumed change in metabolites under freeze stress. Gene family members with no log fold change not presented for brevity

greater levels of damage were observed, refuting the toward greater freeze resistance in grapevine and that the hypothesis that acclimation treatment would decrease current methodology for determining candidate genes freeze damage in grapevine. This result is in contrast to and pathways from cold “hardy” and “sensitive” varieties studies in Arabidopsis32, where low-temperature treat- using low non-freezing temperatures should be reeval- ment primes freeze resistance and leads to greater freeze uated. In fact, this commonly used cold stress treatment survival. However, it should be noted that a similar result seems to uncover the transcriptional response that con- was observed for two grapevine cultivars “Siegrebbe” and tributes to reduced freeze resistance. It is possible that this “Madeleine Angevine” tested at a much earlier phenolo- result partially explains why some studies show that cold gical stage than the cuttings in our study15. This ﬁnding sensitive grapevine cultivars have greater changes in gene suggests that the transcriptional, protein, or metabolic expression when compared with tolerant varieties under response to low temperature is not an adaptive change low-temperature stress21. Londo et al. Horticulture Research (2018) 5:10 Page 10 of 14

Transcriptional Reprogramming in Response to Stress carotene isomerase were observed in acclimation and More than 18 000, genes of the V1 annotation were acclimation freeze treatments while downstream genes detected in this study as being expressed in grapevine leaf were downregulated in freeze, suggesting that there is not tissue. Acclimation treatments resulted in the smallest a concerted recruitment of the early genes leading to ABA number (3 120) of DE genes compared to the control. synthesis. NCED3 was downregulated in acclimation, but Acclimation freeze and freeze treatments resulted in was greatly upregulated in both freeze and acclimation 1.5 × (4 393) and 2 × (5 961) as many DE genes as accli- freeze, and ABA1 was upregulated in freeze shock, driving mation, demonstrating a more dynamic alteration in the ABA synthesis from existing metabolite pools. Continuing transcriptome when leaves are exposed to freezing stres- downstream, freeze treatments increased the expression ses, even when freeze exposure is short (<1 h). of genes responsible for ABA degradation into phaseic We chose to target our analysis of the data set to a acid and subsequently toward the activation of the protein defined subset of biologically relevant pathways and gene phosphatase 2C gene, ABI1. ABI1 was upregulated in families that were consistently observed across treatments freeze, but downregulated in acclimation and acclimation and cultivars. To our knowledge, this is the first evalua- freeze, suggesting that acclimation treatment in some way tion of the grape leaf transcriptome under these stress buffered ABI1 gene expression. Previous studies have conditions and the patterns observed in this data now demonstrated that loss of function of ABA1 and/or ABI1 provide a foundational basis for testing hypotheses of results in lower ABA synthesis and perception, reduced putative metabolites for freeze response in grapevine expression of cold-regulated genes, and also correlates leaves. with loss of freeze resistance in Arabidopsis10, which could explain the increased damage we observed in Ethylene acclimation freeze treatments. Downstream of ABI1, The results of this study repeatedly demonstrated that upregulation of OST and SNRK2 genes as well as upre- pathways related to the synthesis and signaling of ethylene gulation of CPK32 leads to the activation of AREB2 and and ABA were enriched for DE genes, consistent with ABF4 (AREB4) genes indicating that ABA-responsive other studies in grapevine examining the roles of these transcription factors are increased under freeze treat- hormones in abiotic stress response33, 34. The dominant ments. Several repressors of ABA synthesis were differ- contrast observed for the ethylene pathway occurred entially modulated by acclimation vs. freeze, with ABI1 between acclimation and freeze treatments. Acclimation being an example but also AHG3, which was up in freeze treatments resulted in general downregulation of ACC but not significantly changed in acclimation. Acclimation- synthase and oxidase genes, as well as homologs for E8, related and freeze-related gene expression of SOS1, SOS2, the results of which could be reduced production of ABI3, and KEG also demonstrate that acclimation and ethylene. Vitis EIN3 genes were upregulated in acclima- freezing stresses result in very different overall patterns of tion, suggesting that signaling from this portion of the gene expression in ABA synthesis and downstream ABA- ethylene pathway is utilized in acclimation conditions. In dependent gene regulation. contrast, all aspects of ethylene production through ACCase and ACC oxidase, as well as MKK9, MPK3, and Transcription Factor Families EIN3 mechanisms were upregulated in freeze conditions Downstream of the ethylene and ABA pathways lay and EIN3 repression by EBF2 was downregulated. These many different stress-associated transcription factors, taken together suggest that while ethylene signaling may which are in part activated by these hormones including be low in acclimation treatment, there appears to be a the NAC, WRKY, and AP2/ERF transcription factor rapid and upregulated role in response to freeze stress. It families35. The NAC transcription factor gene family has should be noted that studies that have examined the been typically associated with the regulation of plant ethylene pathway in grape exposed to acclimation stress growth and development, as well as biotic and abiotic observed increases in ACCase and ACC oxidase, as well as stress response36. In this study, we detected significant boosted ethylene at earlier time points (8 h post expo- differential expression of 32 of the potential 75 anno- sure), but returning to control levels by 48 h23, which is in tated NAC family members in the Vitisnet/V1 annota- agreement with our results. Taken together, our results tion. The expression patterns of the NAC genes were demonstrate that with or without acclimation pre-expo- quite different between acclimation and freeze treat- sure, the ethylene pathway is rapidly stimulated in leaf ments. In acclimation, 11 NAC members were differ- tissue following exposure to freezing stress conditions. entially expressed, of which 7 were downregulated and 4 were upregulated. A previous study described 8 NAC Abscisic Acid gene family members with differential expression in ABA synthesis and signaling were impacted in a similar previous studies of cold stress, and it is interesting that way to ethylene. Up and downregulation of genes like our results mirror those of these earlier stress studies36. Londo et al. Horticulture Research (2018) 5:10 Page 11 of 14

In contrast, a major change in the transcriptional land- The AP2/ERF transcription factors make up one of the scape occurred following freeze treatments; 28 NAC largest transcription factor families in plants and are members were differentially expressed, and the vast characterized by having at least one AP2 DNA-binding majority represented upregulated genes. Fifteen of these domain. This gene family is both large and highly diverse NAC had similar fold change in freeze and acclimation in gene action, with some genes typically associated with freeze treatments, indicating many NAC genes respond growth and development of cells and plant organs, such as to freeze stress with or without low temperature pre- the AINTEGUMENTA40, 41, SHINE42, and WRINKLED treatment. Of the NAC family members, only VvNAC18 genes43, and other genes which play roles in the percep- and VvNAC26 were seen to be regulated in the same tion of plant hormones and response to biotic and abiotic – manner in all three cold treatments. VvNAC26 was stresses like CBF/DREB and RAP2 genes44 47. The tran- previously observed to have the greatest expression scriptional pattern of AP2/ERF gene expression was change in response to abiotic stress treatments in a highly divergent between acclimation and freeze stress family-wide analysis, and our results support this as a treatments. In acclimation treatment, the vast majority of general COR NAC gene. VvNAC18 was seen to be genes in this family were downregulated in both the inflorescence specific, which does not agree with the growth-related and stress-related modules including Vitis analysis here but it is also important to note the large CBF genes, which have been shown to respond to cold increase of DE NAC genes in response to freeze, stress and regulate downstream cold responses. Three of demonstrating a much wider role for NAC genes in cold the Vitis CBF genes have been shown to be responsive to stress response than previously revealed36. cold, drought, and other stress treatments in previous The WRKY transcription factor gene family was also studies11. However, VvCBF4 has been found to respond enriched in our study, and this family has been associated specifically to acclimation stress (4 °C), with its expression with the regulation of the transcriptome in response to increasing quickly after exposure to acclimation tem- biotic and abiotic stresses and a number of studies have peratures before stabilizing through 8 h of exposure12. examined this family in grapevine24, 37. Many of the VvCBF1 was not detected in our study and VvCBF2 and WRKY genes also seem to play a role regulating the cross- VvCBF3 transcripts were detected but at levels below our talk associated with plant hormone responses to stress, expression cutoff, suggesting that none of these multi- with many of the gene family members responding to stress responsive genes play a major role in grapevine leaf ABA, ethylene, jasmonic acid, and salicylic acid expo- cold stress response. However, it should be noted that sure38. Previous studies of WRKY gene expression Vitis CBF genes may be under different temporal reg- detected by microchip experiments or reverse ulation and could explain the lack of expression noted transcription–PCR in response to cold temperatures have here3. This result also does not agree with previous stu- been conducted using the acclimation temperature used dies11 but could represent variation between the cultivars here and identified a number of cold responsive WRKY or clones used in this study, as has also been shown for genes24, 37. The gene nomenclature39 used in these two cold response of VvCBF2 in the cultivar “Koshu”48. studies do not correspond, but correlating the results of VvCBF4 was greatly downregulated in the acclimation these studies suggest many of the same genes respond to treatment relative to control expression levels (logFC = cold stress. In our study, the expression pattern differ- −5.69). Although this is in disagreement with the peak in ences between acclimation treatment and the two freeze expression seen previously12, our samples were collected treatments largely mirrored that of the previous studies. after 48 h of exposure. Therefore, it appears that early Thirty WRKY genes were differentially expressed in this signaling through the VvCBF4 regulon diminishes over experiment in at least one of the cold temperatures. In time. acclimation treatment, 18 genes were DE, 4 upregulated The activity of the AP2/ERF genes drastically changes in and 14 downregulated. In freeze and acclimation freeze freeze and acclimation freeze treatments. It is under these treatments, the number of genes increased dramatically conditions that we see a clear difference in the gene and the vast majority were seen to be upregulated. Only regulation between AP2/ERF genes governing stress two of the DE genes were shared among all three cold response versus growth and development. The freeze treatments, corresponding with VvWRKY28 and treatment resulted in upregulation and large fold changes VvWRKY21 from previous studies24. Both of these genes in the abiotic stress-associated modules and down- were seen to respond to ABA treatment, suggesting regulation and large fold reductions for growth-related interaction between the ABA results here and WRKY gene modules. This suggests that the perception of imminent expression. It is clear from our study that the transcrip- freezing results in reprogramming of the AP2/ERF gene tional landscape and downstream signaling dependent on family and, consequently, downstream genes. The land- WRKY expression is very different between acclimation scape was also similar in both freezing conditions, though and freeze treatments. acclimation freeze treatments represent a subset of the Londo et al. Horticulture Research (2018) 5:10 Page 12 of 14

total freeze response genes and the logFC of these genes expression for several aspects of these pathways were are generally lower than those in freeze. These results shared. All cold treatments resulted in the upregulation of suggest that the rapid and dynamic changes that occur enzymes of starch synthesis compared to the control, when freeze shock conditions occur, are in some ways suggesting a general cold stress strategy of increasing prevented or not needed if grapevine is exposed to lower non-soluble sugar reserves and this is consistent with temperatures prior to freezing conditions. Upregulation of other cold stress studies where starch levels were mea- VvCBF4 was observed in both freeze and acclimation sured after cold (4 °C) exposure53. Interestingly, genes that freeze conditions, though the changes were only sig- encode enzymes responsible for the degradation of starch nificant between control treatment and acclimation freeze into maltose and dextrin were also upregulated, suggest- treatment (logFC = 1.61). This result is interesting as it ing starch may not be the terminal response to cold. demonstrates a potentially nuanced regulation of VvCBF4 Genes responsible for the production of sucrose were also by varied cold exposure. As mentioned above12, VvCBF4 upregulated in all cold treatments demonstrating that typically increases initially when exposed to acclimation sugar reserves may be partitioned between non-soluble temperatures, but then decreases to control levels or and soluble sugar forms. The relative levels of gene lower as seen in this study, by 48 h. We show that VvCBF4 expression change tended to demonstrate that freeze expression increases as a result of freeze shock, but being shock stress resulted in greater downregulation than that pre-treated with acclimation temperature results in a of acclimation and acclimation freeze. Taken together, all much higher, significant upregulation of this master cold stresses had the effect of downregulating the pro- controller. The reason for this primed effect are not clear, duction of simple soluble sugars that are used in general but we can speculate that other aspects of gene signaling metabolism, likely resulting in the reduction in growth that occur upstream of VvCBF4 may be also primed and and development that often occurs at cold temperatures. could enable greater activation of this cold regulon should In contrast to these shared DE patterns, a few key freeze conditions occur. It is clear from these results that changes in gene expression suggests that these sugar stress response and gene regulation of the AP2/ERF family pathways are utilized differently between acclimation and is drastically different between acclimation and freeze freeze stresses. In both freeze treatments, the expression treatments. Many of the genes identified as upregulated in pattern of phenol-β-glucosyltransferase was upregulated. both freeze and acclimation freeze represent new candi- This enzyme catalyzes the hydrolysis of UDP-glucose to date genes that can be investigated for potential use to D-glucoside, a general glucose based sugar backbone used increase the freeze resistance in grapevine. in a multitude of pathways related to secondary product biosynthesis, such as flavanols, terpenes and stilbenes. Sugar and Starch Upregulation of the portion of the pathway leading to the Downstream of plant hormone signaling and the tran- production of raffinose was also seen in response to freeze scription factors that respond to stress are a multitude of treatment, in particular the committing step of raffinose metabolism and biosynthesis pathways that play a role in synthesis. While we did not directly measure raffinose modifying plant physiology to resist or tolerate cold stress. concentrations in this study, it is clear from our data that Sugars have long been associated with increases in chill freeze induces all the genes leading from simple soluble and freeze resistance in a number of crops species49, 50. sugars to raffinose. Increased raffinose concentrations and Soluble sugars like sucrose, glucose, fructose, and raffi- production has been seen to be a common cold stress nose have been shown to increase in concentration when response in a number of different plant systems including plant tissues are cold stressed. These sugars are thought both perennials and annual species54, 55. However, it has to provide freeze protection through stabilization of also been seen that raffinose itself is not necessary for membranes, scavenging of reactive oxygen species, acting freezing tolerance as mutant Arabidopsis lines with lack as signaling molecules, and by decreasing the freezing of raffinose synthase are still able to survive freezing point as compatible osmolytes51, 52. conditions56. Raffinose could have many different cryo- Studies conducted to measure the carbohydrate fluxes protective roles, ranging from ROS scavenging to stabili- of grapevine leaves exposed to acclimation stress have zation of photosynthetic proteins57. In grapevine, shown that these sugars do increase over time with glu- increased levels of soluble sugars, including raffinose, are cose, fructose, and sucrose increasing initially and raffi- correlated with changes in dormant tissue cold hardi- nose and galactinol increasing after several days of ness58 but no direct mechanism has been uncovered. exposure53. We observed that DE genes were enriched for Treatment of grapevine leaves to moderate chill stress of the sucrose and starch metabolism and galactose path- 15/7 °C day/night temperatures increased leaf raffinose ways. Unlike the contrasting effects of acclimating tem- concentrations, and that more “hardy” varieties accumu- peratures versus freeze temperatures observed for ABA, lated a higher concentration59. All data to date are cor- ethylene, and transcription factors, the pattern of gene relative suggesting that raffinose concentration is tied to Londo et al. Horticulture Research (2018) 5:10 Page 13 of 14

cold stress resistance, and this study further supports a examination of grapevine cold stress response at 4 °C role for raffinose in cold resistance. It is clear from our many not fully elucidate or identify the most important data that there are fundamental differences in the gene aspects of freeze response. Our results demonstrate that regulation and signaling of the carbohydrate pools the transcriptional landscape of cold stress at 4 °C and between acclimated and freeze shock exposed grapevine that of freezing stress at −3 °C are very different in key leaves. hormone, transcription factor, and sugar pathways As with any assessment of transcriptomic response, future Cultivar responses studies are needed to validate the changes in gene Our experiment clearly demonstrated that low tem- expression seen here with metabolomic profiling as well perature pretreatment of grapevine cuttings does not as further investigation into the potential role of stilbenes increase resistance to freezing stress. The phenotypic in freeze response. range of freeze damage was variable between cultivars and Acknowledgements similarly, preventing a simplistic assessment of what All research was supported by USDA appropriated funds for project number unique aspects of gene regulation contributes to increased 1910-21220-006-00D. We would like to acknowledge partial support for A.P.K., freeze resistance. The number of significant DEs in each by CAPES, Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, Brazil. We would like to acknowledge and thank Kathleen Deys of USDA-ARS- cultivar did not correlate with observed leaf damage. Both GGRU for contributing to the field collection and processing of leaf tissues “Cabernet Franc” and “Sangiovese” had a high number of used in this study. We would like to acknowledge Dr. Anne Fennell, South DE genes in all treatment contrasts yet represent very Dakota State University, for contributing to analysis and manuscript different phenotypes. As “Sangiovese” seemed resistant to preparation. freeze damage both in freeze and acclimation freeze Author details treatments, we explored which genes and pathways were 1United States Department of Agriculture, Agricultural Research Service, Grape 2 upregulated, enriched, and unique for “Sangiovese”.Of Genetics Research Unit, 630 W. North Street, Geneva, NY, USA. School of Integrative Plant Science, Horticulture section, Cornell University-New York particular note was the phenlypropanoid pathway, and State Agricultural Experiment Station, 630 W. North Street, Geneva, NY, USA. important pathway in secondary metabolism. Stilbenes 3Genomics Research Center, University of Rochester Medical Center, Rochester, are a class of plant protective gene products called phy- NY, USA toalexins and have been shown to be important for plants fl 60 Con icts of Interest exposed to a multitude of abiotic and biotic stresses . The authors declare that they have no conflict of interest. While a few stilbene synthesis genes were upregulated in “Cabernet Franc” (9), “Chardonnay” (8), and “Riesling” (7) Supplementary Information accompanies this paper at https://doi.org/ 10.1038/s41438-018-0020-7. in freeze or acclimation freeze stresses, 24 of 42 stilbene “ ” synthase genes were upregulated in the Sangiovese Received: 22 September 2017 Revised: 2 January 2018 Accepted: 11 samples and 7 were uniquely found in “Sangiovese”. Stil- January 2018 bene induction under cold temperature exposure has not previously been investigated and the transcription factors responsible for stilbene synthase expression are unre- solved. However, stilbenes were seen to be induced by References 60 1. Wisniewski,M.,Bassett,C.&Gusta,L.V.Anoverviewofcoldhardinessin other stresses that induce ABA and ethylene pathways woody plants: seeing the forest through the trees. HortScience 38,952–959 and stimulation here may be a result of the changes seen (2003). in these hormone pathways. Unfortunately, it is not pos- 2. Gusta, L. V. & Wisniewski, M. Understanding plant cold hardiness: an opinion. Physiol. Plant. 147,4–14 (2013). sible to determine if stilbenes are the reason for increased 3. Wisniewski, M. et al. Genomics of cold hardiness in woody plants. Crit. Rev. freeze resistance in “Sangiovese” or if it simply has greater Plant. Sci. 33,92–124 (2014). basal expression of these genes, but this suggests that the 4. Mizoguchi, T. et al. A gene encoding a mitogen-activated protein kinase kinase kinase is induced simultaneously with genes for a mitogen-activated role of stilbenes in freeze response may be worth further protein kinase and an S6 ribosomal protein kinase by touch, cold, and water study. stress in Arabidopsis thaliana. Proc. Natl Acad. Sci. USA 93, 765–769 (1996). 5. Polisenski, D. H. & Braam, J. Cold shock regulation of the Arabidopsis TCH genes and effects of modulating intracellular calcium levels. Plant Physiol. 111, Conclusion 1271–1279 (1996). Young grapevine tissues are faced with the possibility of 6. Yamaguchi-Shinozaki, K. & Shinozaki, K. A novel cis-acting element in an freezing stress as a result of frost events and resistance to Arabidopsis gene is involved in responsiveness to drought, low temperature, or high-salt stress. Plant Cell 6,251–264 (1994). this stress is important to prevent damage in the vineyard. 7. Stockinger, E. J., Glimour, S. J. & Thomashow, M. F. Arabidopsis thaliana CBF1 Our study showed that cultivated grapevine appears to encodes an AP2 domain-containing transcription activator that binds to the have no acclimation ability in green tissues such that pre- C-repeat/DRE, a cis-acting DNA regulatory element that stimulates transcription in response to low temperature and water deficit. Proc. Natl Acad. Sci. USA treatment of leaves to low, non-freezing temperatures (4 ° 94,1035–1040 (1997). C) increased damage when leaves were exposed to a short 8. Chinnusamy, V. et al. ICE1: a regulator of cold-induced transcriptome and duration freeze treatment. Our study indicates that freezing tolerance in Arabidopsis. Genes Dev. 15,912–924 (2003). Londo et al. Horticulture Research (2018) 5:10 Page 14 of 14

9. Jaglo-Ottosen, K. R., Glimour, S. J., Zarka, D. G., Schabenberger, O. & Thoma- 36. Wang, N., Zheng, Y., Xin, H., Fang, L. & Li, S. Comprehensive analysis of NAC show, M. F. Arabidopsis CBF1 overexpression induces COR genes and domain transcription factor gene family in Vitis vinifera. Plant Cell Rep. 32, enhances freezing tolerance. Science 280,104–106 (1998). 61–75 (2013). 10. Thomashow, M. F. Plant cold acclimation: freezing tolerance genes and reg- 37. Wang, L. et al. Genome-wide identification of WRKY family genes and their ulatory mechanisms. Annu. Rev. Plant Physiol. Plant Mol. Biol. 50,571–599 response to cold stress in Vitis vinifera. BMC Plant Biol. 14, 103 (2014). (1999). 38. Guo, C. et al. Evolution and expression analysis of the grape (Vitis vinifera L.) 11. Xiao, H., Siddiqua, M., Braybrook, S. & Nassuth, A. Three grape CBF/DREB1 WRKY gene family. J. Exp. Bot. 65,1513–1528 (2014). genes respond to low temperature, drought and abscisic acid. Plant Cell 39. Grimplet, J. et al. The grapevine gene nomenclature system. BMC Genomics Environ. 29, 1410–1421 (2006). 15, 1077 (2014). 12. Xiao,H.,Tattersall,E.A.R.,Siddiqua,M.K.,Cramer,G.R.&Nassuth,A.CBF4isa 40. Mizukami, Y. & Fischer, R. L. Plant organ size control: AINTEGUMENTA regulates unique member of the CBF transcription factor family of Vitis vinifera and Vitis growth and cell numbers during organogenesis. Proc. Natl Acad. Sci. USA 97, riparia. Plant Cell Environ. 31,1–10 (2008). 942–947 (1999). 13. Rahman M. A., Moody M. A., Nassuth A. Grape contains 4 ICE genes whose 41. Chialva, C. et al. Expression of grapevine AINTEGUMENTA-like genes is asso- expression includes alternative polyadenylation, leading to transcripts ciated with variation in ovary and berry size. Plant Mol. Biol. 91,67–80 (2016). encoding at least 7 different ICE proteins. Environ. Exp. Bot. 106,70–78 (2014). 42.Shi,J.X.etal.SHINEtranscriptionfactorsactredundantlytopatternthe https://doi.org/10.1016/j.envexpbot.2014.01.003. archetypal surface of Arabidopsis flower organs. PLoS Genet. 7, e1001388 14. Mahajan, S. & Tuteja, N. Cold, salinity and drought stresses: an overview. Arch. (2011). Biochem. Biophys. 444,139–158 (2005). 43. To, A. et al. WRINKLED Transcription factors orchestrate tissue-specficregula- 15. Fuller, M. P. & Telli, G. An investigation of the frost hardiness of grapevine (Vitis tion of fatty acid biosynthesis in arabidopsis. Plant Cell 24, 5007–5023 (2012). vinifera)duringbudbreak.Ann. Appl. Biol. 135,589–595 (1999). 44. Lin,R.-C.,Park,H.-J.&Wang,H.-Y.Roleof arabidopsis RAP2.4 in regulating light- 16. Wisniewski, M., Glenn, D. M. & Fuller, M. P. Use of a hydrophobic particle film as and ethylene- mediated developmental processes and drought stress toler- a barrier to extrinsic ice nucleation in tomato plants. J. Am. Soc. Hortic. 127, ance. Mol. Plant 1,42–57 (2008). 358–364 (2002). 45. Dong, C.-J. & Liu, J.-Y. The arabidopsis EAR-motif-containing protein RAP2.1 17. Mills, L. J., Ferguson, J. C. & Keller, M. Cold-hardiness evaluation of grapevine functions as an active transcriptional repressor to keep stress responses under buds and cane tissues. Am.J.Enol.Vitic.55,194–200 (2006). tight control. BMC Plant Biol. 10,47(2010). 18. Londo J. P., Kovaleski A. P. Characterization of wild north american grapevine 46. Zhao, Y. et al. Arabidopsis RAP2.2 plays an important role in plant resistance to cold hardiness using differential thermal analysis. Am.J.Enol.Vitic. 2017. Botrytis cinerea and ethylene responses. New Phytol. 195, 450–460 (2012). https://doi.org/10.5344/ajev.2016.16090 47. Zhao, T., Xia, H., Liu, J. & Ma, F. The gene family of dehydration responsive 19. Tattersall, E. A. R. et al. Transcript abundance profiles reveal larger and more element-binding transcription factors in grape (Vitis vinifera): genome-wide complex responses of grapevine to chilling compared to osmotic and salinity identification and analysis, expression profiles, and involvement in abiotic stress. Functional and Integrated. Genomics 7,317–333 (2007). stress resistance. Mol. Biol. Rep. 41,1577–1590 (2014). Genetics 7(5):e1001388. 20. Yang, Y. et al. Identification of the dehydrin gene family from grapevine 48. Takuhara Y., Kobayashi M., Suzuki S. Low-temperature-induced transcription species and analysis of their responsivenesstovariousformsofabioticand factors in grapevine enhance cold tolerance in transgenic Arabidopsis plants. biotic stress. BMC Plant Biol. 12, 140 (2012). Journal of Plant Physiology. 168,967–975 (2011). 21. Xin, H. et al. Genome wide transcriptional profile analysis of Vitis amurensis 49. Gusta, L. V., Wisniewski, M., Nesbitt, N. T. & Gusta, M. L. The effect of water, and Vitis vinifera in response to cold stress. PLoS ONE 8, e58740 (2013). sugars, and proteins on the pattern of ice nucleation and propagation in 22. Xu, W. et al. Transcriptome profiling of Vitis amurensis, an extremely cold- acclimated and non-acclimated canola leaves. Plant Physiol. 135,1642–1653 tolerant Chinese wild Vitis species, reveals candidate genes and events that (2004). potentially connected to cold stress. Plant Mol. Biol. 86,527–541 (2014). 50. Theocharis, A., Clement, C. & Ait Barka, E. Physiological and molecular changes 23. Sun, X. et al. Ethylene positively regulates cold tolerance in grapevine by in plants grown at low temperatures. Planta 235,1091–1105 (2012). modulating the expression of ETHYLENE RESPONSE FACTOR 057. Sci. Rep. 6, 51. Nishizawa,A.,Yabuta,Y.&Shigeoka,S.Galactinolandraffinose constitute a 24066 (2016). novel function to protect plants from oxidative damage. Plant Physiol. 147, 24. Wang M. et al. Genome and transcriptome analysis of the grapevine (Vitis 1251–1263 (2008). vinifera L.) WRKY gene family. Hortic. Res. 1, 14016 (2014). https://doi.org/ 52. Ruelland E. et al.inAdvances in Botanical Research.(edsJ.-C.Kader&Delseny 10.1038/hortres.2014.16 M.) (Academic Press, London, 2009). 25. Londo,J.P.&Johnson,L.M.Variation in the chilling requirement and bud 53. Fernandez, O. et al Burkholderia phytofirmans PsJN acclimates grapevine to burst rate of wild Vitis species. Environ. Exp. Bot. 160,138–147 (2014). cold by modulating carbohydrate metabolism. Mol. Plant Microbe Interact. 25, 26. Coombe, B. G. Growth stages of the grapevine: adoption of a system for 496–504 (2012). identifying grapevine growth stages. Aust. J. Grape Wine Res. 1,104–110 (1995). 54. Strimbeck,G.R.,Kjellsen,T.D.,Schaberg,P.G.&Murakami,P.F.Coldinthe 27. Borodina, T., Adjaye, J. & Sultan, M. Chapter five—astrand-specific library common garden: comparative low-temperature tolerance of boreal and preparation protocol for RNA sequencing. Methods Enzymol. 500,79–98 temperate conifer foliage. Trees Struct. Funct. 21,557–567 (2007). (2011). 55. Klotke,J.,Kopka,J.,Gatzke,N.&Heyer,A.G.Impactofsolublesugarcon- 28.Osier,M.V.VitisPathways:genepathwayanalysisforV.vinifera.Vitis 55, centrations on the acquisition of freezing tolerance in accessions of Arabi- 129–133 (2016). dopsis thaliana with contrastingcoldadaptation—evidence for a role of 29. Grimplet, J. et al. VitisNet: “Omics” integration through grapevine molecular raffinose in cold acclimation. Plant Cell Environ. 27,1395–1404 (2004). networks. PLoS ONE 4, e8365 (2009). 56. Zuther, E. et al. The role of raffinose in the cold acclimation response of 30. Heberle, H., Meirelles, G. V., da Silva, F. R., Telles, G. P. & Minghim, R. Inter- Arabidopsis thaliana. FEBS Lett. 576,169–173 (2004). actiVenn: a web-based tool for the analysis of sets through Venn diagrams. 57. Knaupp,M.,Mishra,K.B.,Nedbal,L.&Heyer,A.G.Evidenceforaroleof BMC Bioinformatics 16, 169 (2015). raffinose in stabilizing photosystem II during freeze-thaw cycles. Planta 234, 31. Xiong, L. & Zhu, J.-K. Regulation of abscisic acid biosynthesis. Plant Physiol. 133, 477–486 (2011). 29–36 (2003). 58. Hamman, R. A., Dami, I.-E. & Walsh, T. M., . & StushnoffC.. Seasonal carbohydrate 32. Wanner, L. A. & Junttila, O. Cold-induced freezing tolerance in arabidopsis. changes and cold hardiness of chardonnay and riesling grapevines. Am. J. environmental and stress. Physiology 120,391–399 (1999). Enol. Vitic. 47,31–36 (1996). 33. Cramer,G.R.,Urano,K.,Delrot,S.,Pezzotti,M.&Shinozaki,K.Effectsofabiotic 59. Grant,T.N.,Dami,I.E.,Ji,T.,Scurlock,D.&Streeter,J.Variationinleafandbud stress on plants: a systems biology perspective. BMC Plant Biol. 11, 163 (2011). soluble sugar concentration among Vitis genotypes grown under two tem- 34. Hopper, D. W., Ghan, R., Schlauch, K. A. & Cramer, G. R. Transcriptomic network perature regimes. Can. J. Plant Sci. 89,961–968 (2009). analyses of leaf dehydration responses identify highly connected ABA and 60. Vannozzi,A.,Dry,I.B.,Fasoli, M., Zenoni, S. & Lucchin, M. Genome-wide analysis ethylene signaling hubs in three grapevine species differing in drought tol- of the grapevine stilbene synthase multigenic family: genomic organization erance. BMC Plant Biol. 16, 118 (2016). and expression profiles upon biotic and abiotic stresses. BMC Plant. Biol. 12, 35. Licausi, F. et al. Genomic and transcriptomic analysis of the AP2/ERF super- 130 (2012). family in Vitis vinifera. BMC Genomics 11, 719 (2010).